One of the things that gave me the confidence to write my bloated new novel Madness is Better than Defeat was finding out that there are eighteen point-of-view characters in George R. R. Martin’s A Dance with Dragons. Eighteen is a remarkable number, but nobody seems to complain that Martin’s books are difficult to follow (I have yet to read any of them myself). I was planning to have a mere eight point-of-view characters in my book, so I reasoned that even if my writing was only 45% as involving as Martin’s, I could keep my readers on board. Of course, given the disparity in our commercial success so far, it seems likely that my writing is not even close to 45% as involving as Martin’s, but at least my approach was feasible in theory.
Nevertheless, I knew I would have to do everything I could to keep my personnel distinct in the reader’s mind. The first way I set about doing this was with their names. Of those eight point-of- view characters, two are more like supporting cast, leaving six real protagonists. I assigned each of them their own dominant vowel: A (Vansaska), E (Whelt), I (Trimble), O (Coehorn), U (Burlingame), and … well, obviously at that point I had run out of vowels, so I called the final character Zonulet, a name with no dominant vowel, intended to symbolise that he was a maverick who intersected with all of the other characters’ lives at different points.
Meanwhile, with the second and tertiary characters, I tried to use names to signal to the reader whether or not they were worth remembering: so if a character was called Sandy Mitchell or Janet Jones, they weren’t going to play an ongoing role in the story, but if they were called Gracie Calix or Colby Droulhiole, they were very likely to pop up again later. (Does George R. R. Martin have minor characters called things like “Murphy” or “Rodriguez”? Again, I haven’t read those books myself, but I assume that he does. It would be very surprising if a writer of such expansive books didn’t employ this useful technique.)
In the past I’ve occasionally inspired a bit of weariness in critics with my use of unusual, “Dickensian” monikers. And indeed many novelists do seem to limit themselves to the most common, inconspicuous names, as if that’s more “realistic.” But just look at – to take a random example that I came across rececntly – the Goldman Sachs new partner list for 2016. It includes people called Court Golumbic, Canute Dalmasse, Padideh Raphael, Aurora Swithenbank and Harold Hope III. People in real life have amazing names (even if some allowance must be made for the rather rarefied demographic of Goldman Sachs partners).
What I’ve never had the courage to do is write two characters with the same name, like the various Lintons, Cathys and Heathcliffs in Wuthering Heights, or the two Jason Compsons in The Sound and the Fury (which has fifteen point-of-view characters, by the way). A novel with a cast as large as mine should really have at least one name duplication, not only for statistical veracity but also because it’s the logical endpoint of the method I’ve described.
In my next book, perhaps the important characters will have both grotesquely exaggerated physical characteristics and names like Benzodiazepine or Sphygmomanometer, while every single one of the unimportant characters will be hairless clones called Mel.